Irreversible Electroporation: Maximizing Treatment Efficacy through Optimization Strategies and Robotic Applications
Girindra Wardhana is a PhD student in the department Robotics and Mechatronics. (Co)Supervisors are prof.dr. J.J. Fütterer and dr.ir. M. Abayazid from the faculty of Electrical Engineering, Mathematics and Computer Science.
Irreversible electroporation is a non-thermal ablation technique that uses an electric field to destroy tumor cells. Compared to thermal ablation methods, irreversible electroporation causes cell death mainly through apoptosis, preserves surrounding tissue, and is not affected by local blood flow. However, because of the varying properties of target tissues, the optimal parameters for irreversible electroporation treatment are still under investigation. To address this challenge, various studies have developed computational models for finding optimal irreversible electroporation protocols that cover the entire tumor region while avoiding unwanted thermal injury.
Electrode placement during irreversible electroporation treatment is a significant challenge for clinicians, with the number of electrodes utilized ranging from two to six, and the distance between each of the electrode pair preferably between 10 and 25 mm. To aid with this task, robotic applications are becoming increasingly popular in assisting clinicians with electrode placement in irreversible electroporation treatment. Studies have shown that robots can reduce intervention time, minimize radiation exposure, and achieve equal or even higher accuracy compared to manual placement.
This thesis presents computational models to investigate the optimal irreversible electroporation protocol by examining the effects of pulse parameters and electrode configurations on the ablation area and thermal damage to the target tissue. Model validation has been conducted on animal and vegetable tissue, with in-depth discussions offered in Chapters 2 and 3. To improve the accuracy of the model, the actual shape of the tissue should be considered during the simulation process. Therefore, an automatic segmentation method using deep learning was proposed to segment liver and tumors from CT images, offering high accuracy and faster processing times compared to manual segmentation, as elaborated in Chapter 4.
Robotic systems have been developed to assist clinicians with the placement of multiple electrodes in target tissues. Two robot designs were proposed and fabricated using 3D printing with plastic material. They were actuated using a pneumatic system, making them suitable for use within an MRI scanner. The initial robot design had four degrees of freedom, and multiple electrode insertions were carried out sequentially while taking into account the orientation of previous electrodes to maintain parallelism between them. The second robot design is an improvement on the first robot, with a grid system employed to accommodate the insertion of multiple electrodes simultaneously while maintaining the distance and parallelism between electrodes. This design has two degrees of freedom, which simplifies robot control and installation by requiring fewer pneumatic tubes. Chapters 5 and 6 present the designs and evaluations of the robot performances. Both designs have demonstrated limited deviation within the acceptable range for irreversible electroporation procedures. The robotic systems have great potential to assist clinicians in accurately placing multiple electrodes into target tissues, ultimately improving treatment outcomes.
In conclusion, the tools proposed in this thesis, including computational models and robotic systems, contribute to the improvement of treatment efficacy using irreversible electroporation. These tools can aid clinicians in developing more patient-specific models, improving treatment planning outcomes, and assisting in electrode placement accuracy.
